Clostridium is a genus of anaerobe gram-positive bacteria, belonging to the Firmicutes. Clostridium consists of around 100 species that include common free-living bacteria as well as important pathogens, such as Clostridium botulinum and Clostridium tetani. Both species produce neurotoxins, botulinum toxin and tetanus toxin, respectively. These neurotoxins are potent inhibitors of calcium-dependent neurotransmitter secretion of neuronal cells and are among the strongest toxins known to man. The lethal dose in humans lies between 0.1 ng and 1 ng per kilogram of body weight.
Oral ingestion of botulinum toxin via contaminated food or generation of botulinum toxin in wounds can cause botulism, which is characterized by paralysis of various muscles. Paralysis of the breathing muscles can cause death of the affected individual.
Although both botulinum neurotoxin (BoNT) and tetanus neurotoxin (TxNT) function via a similar initial physiological mechanism of action, inhibiting neurotransmitter release from the axon of the affected neuron into the synapse, they differ in their clinical response. While the botulinum toxin acts at the neuromuscular junction and other cholinergic synapses in the peripheral nervous system, inhibiting the release of the neurotransmitter acetylcholine and thereby causing flaccid paralysis, the tetanus toxin acts mainly in the central nervous system, preventing the release of the inhibitory neurotransmitters GABA (gamma-aminobutyric acid) and glycine by degrading the protein synaptobrevin. The consequent overactivity in the muscles results in generalized contractions of the agonist and antagonist musculature, termed a tetanic spasm (rigid paralysis).
While the tetanus neurotoxin exists in one immunologically distinct type, the botulinum neurotoxins are known to occur in seven different immunogenic types, termed BoNT/A through BoNT/H. Most Clostridium botulinum strains produce one type of neurotoxin, but strains producing multiple toxins have also been described.
Botulinum and tetanus neurotoxins have highly homologous amino acid sequences and show a similar domain structure. Their biologically active form comprises two peptide chains, a light chain of about 50 kDa and a heavy chain of about 100 kDa, linked by a disulfide bond. A linker or loop region, whose length varies among different clostridial toxins, is located between the two cysteine residues forming the disulfide bond. This loop region is proteolytically cleaved by an unknown clostridial endoprotease to obtain the biologically active toxin.
The molecular mechanism of intoxication by TxNT and BoNT appears to be similar as well: entry into the target neuron is mediated by binding of the C-terminal part of the heavy chain to a specific cell surface receptor; the toxin is then taken up by receptor-mediated endocytosis. The low pH in the so formed endosome then triggers a conformational change in the clostridial toxin which allows it to embed itself in the endosomal membrane and to translocate through the endosomal membrane into the cytoplasm, where the disulfide bond joining the heavy and the light chain is reduced. The light chain can then selectively cleave so called SNARE-proteins, which are essential for different steps of neurotransmitter release into the synaptic cleft, e.g. recognition, docking and fusion of neurotransmitter-containing vesicles with the plasma membrane. TxNT, BoNT/B, BoNT/D, BoNT/F, and BoNT/G cause proteolytic cleavage of synaptobrevin or VAMP (vesicle-associated membrane protein), BoNT/A and BoNT/E cleave the plasma membrane-associated protein SNAP-25, and BoNT/C cleaves the integral plasma membrane protein syntaxin and SNAP-25.
Clostridial neurotoxins display variable durations of action that are serotype specific. The clinical therapeutic effect of BoNT/A lasts approximately 3 months for neuromuscular disorders and 6 to 12 months for hyperhidrosis. The effects of BoNT/E, on the other hand, last less than 4 weeks. The longer lasting therapeutic effect of BoNT/A makes it preferable for clinical use compared to the other serotypes, for example serotypes B, C1, D, E, F, G and H. One possible explanation for the divergent durations of action might be the distinct subcellular localizations of BoNT serotypes. The protease domain of BoNT/A light chain localizes in a punctate manner to the plasma membrane of neuronal cells, co-localizing with its substrate SNAP-25. In contrast, the short-duration BoNT/E serotype is cytoplasmic. Membrane association might protect BoNT/A from cytosolic degradation mechanisms allowing for prolonged persistence of BoNT/A in the neuronal cell.
In Clostridium botulinum, the botulinum toxin is formed as a protein complex comprising the neurotoxic component and non-toxic proteins. The accessory proteins embed the neurotoxic component thereby protecting it from degradation by digestive enzymes in the gastrointestinal tract. Thus, botulinum neurotoxins of most serotypes are orally toxic. Complexes with, for example, 450 kDa or with 900 kDa are obtainable from cultures of Clostridium botulinum. 
In recent years, botulinum neurotoxins have been used as therapeutic agents in the treatment of dystonias and spasms. Preparations comprising botulinum toxin complexes are commercially available, e.g. from Ipsen Ltd (Dysport®) or Allergan Inc. (Botox®). A high purity neurotoxic component, free of any complexing proteins, is for example available from Merz Pharmaceuticals GmbH, Frankfurt (Xeomin®).
Clostridial neurotoxins are usually injected into the affected muscle tissue, bringing the agent close to the neuromuscular end plate, i.e. close to the cellular receptor mediating its uptake into the nerve cell controlling said affected muscle. Various degrees of neurotoxin spread have been observed. The neurotoxin spread is thought to depend on the injected amount and the particular neurotoxin preparation. It can result in adverse side effects such as paralysis in nearby muscle tissue, which can largely be avoided by reducing the injected doses to the therapeutically relevant level. Overdosing can also trigger the immune system to generate neutralizing antibodies that inactivate the neurotoxin preventing it from relieving the involuntary muscle activity. Immunologic tolerance to botulinum toxin has been shown to correlate with cumulative doses.
At present, clostridial neurotoxins are still predominantly produced by fermentation processes using appropriate Clostridium strains. However, industrial production of clostridial neurotoxin from anaerobic Clostridium culture is a cumbersome and time-consuming process. Due to the high toxicity of the final product, the procedure must be performed under strict containment. During the fermentation process, the single-chain precursors are proteolytically cleaved by an unknown clostridial protease to obtain the biologically active di-chain clostridial neurotoxin. The degree of neurotoxin activation by proteolytic cleavage varies between different strains and neurotoxin serotypes, which is a major consideration for the manufacture due to the requirement of neurotoxin preparations with a well-defined biological activity. Furthermore, during fermentation processes using Clostridium strains the clostridial neurotoxins are produced as protein complexes, in which the neurotoxic component is embedded by accessory proteins. These accessory proteins have no beneficial effect on biological activity or duration of effect. They can however trigger an immune reaction in the patient, resulting in immunity against the clostridial neurotoxin. Manufacture of recombinant clostridial neurotoxins, which are not embedded by auxiliary proteins, might therefore be advantageous.
Methods for the recombinant expression of clostridial neurotoxins in E. coli are well known in the art (see, for example, WO 00/12728, WO 01/14570, or WO 2006/076902). Furthermore, clostridial neurotoxins have been expressed in eukaryotic expression systems, such as in Pichia pastoris, Pichia methanolica, Saccharomyces cerevisiae, insect cells and mammalian cells (see WO 2006/017749).
Recombinant clostridial neurotoxins may be expressed as single-chain precursors, which subsequently have to be proteolytically cleaved to obtain the final biologically active clostridial neurotoxin. Thus, clostridial neurotoxins may be expressed in high yield in rapidly-growing bacteria as relatively non-toxic single-chain polypeptides.
Furthermore, it might be advantageous to modify clostridial neurotoxin characteristics regarding biological activity, cell specificity, antigenic potential and duration of effect by genetic engineering to obtain recombinant neurotoxins with new therapeutic properties in specific clinical areas. Genetic modification of clostridial neurotoxins might allow altering the mode of action or expanding the range of therapeutic targets.
WO 96/39166 discloses analogues of botulinum toxin comprising amino acid residues which are more resistant to degradation in neuromuscular tissue.
Patent family based on WO 02/08268 (including family member U.S. Pat. No. 6,903,187) discloses a clostridial neurotoxin comprising a structural modification selected from addition or deletion of a leucine-based motif, which alters the biological persistence of the neurotoxin (see also: Fernández-Salas et al., Proc. Natl. Acad. Sci. U.S.A. 101 (2004) 3208-3213; Wang et al., J. Biol. Chem. 286 (2011) 6375-6385). Fernández-Salas et al. initially hypothesized that the increased persistence was due to the membrane-binding properties of the dileucine motif (see Fernández-Salas et al., loc. cit., p. 3211 and 3213). Wang et al. mention this membrane theory (see Wang et al., loc. cit., p. 6376, left column, last full paragraph, and p. 6383, first full paragraph of “Discussion”), but favor an alternative theory: the protection from degradation by proteolysis (see Wang et al., loc. cit., p. 6384, left column, lines 27ff).
US 2002/0127247 describes clostridial neurotoxins comprising modifications in secondary modification sites and exhibiting altered biological persistence.
Botulinum toxin variants exhibiting longer biological half lives in neuromuscular tissue than naturally occurring botulinum toxins would be advantageous in order to reduce administration frequency and the incidence of neutralizing antibody generation since immunologic tolerance to botulinum toxin is correlated with cumulative doses.
Furthermore, BoNT serotypes naturally exhibiting a short duration of action could potentially be effectively used in clinical applications, if their biological persistence could be enhanced. Modified BoNT/E with an increased duration of action could potentially be used in patients exhibiting an immune reaction against BoNT/A. Moreover, BoNT/E was shown to induce a more severe block of pain mediator release from sensory neurons than BoNT/A. In clinical applications where BoNT/A provides only partial pain relief or in just a subset of patients, such as in the treatment of headaches, or where BoNT/E has been found to be more effective than BoNT/A but gives only short-term therapy, such as in the treatment of epilepsy, BoNT/E with an increased duration of effect might prove useful.
There is a strong demand to produce clostridial neurotoxins with an increased duration of effect, in order to allow for reduction of administration frequency and exploitation of the therapeutic potential of BoNT serotypes, which have so far been considered impractical for clinical application due to the short half-life of the respective clinically relevant effect. Ideally, the duration of effect of a particular clostridial neurotoxin could be adjusted in a tailor-made fashion in order to address any particular features and demands of a given indication, such as the amount of neurotoxin being administered, frequency of administration etc. To date, such aspects have not been solved satisfactorily.